Friendly fire losses have always been a problem, even in the days before guns. But during World War II, it was later (using interviews of veterans) found that some 20 percent of U.S. casualties were probably from friendly fire. The official friendly fire rate for World War II was 1.5 percent. The combat veterans were particularly reluctant to talk about instances where rifle fire or grenades were the cause of friendly fire losses. But this was quite common, and confirmed by checking with battlefield surgeons (who could tell a U.S. bullet from an enemy one.) The percentage apparently stayed the same in Korea, but may have gone up a bit during the Vietnam war (where the official friendly fire rate was 2.85 percent.) The most common place for friendly fire casualties to occur, and not be reported as such, is during infantry combat. In fire fights, or battle in urban areas, gun fire and grenades were going every which way and it was often hard to determine who was hit by who. The only witnesses to this sort of the thing were reluctant to report it. For one thing, everyone knew that any of them could be the one to shoot one of their friends in the chaos of combat. Having your friends killed in battle was bad enough without having the dead man's family know it was American troops that did it.

Friendly fire from your own artillery and air craft was always easier to spot and report. This can be seen from the pattern of officially reported friendly fire casualties from past wars. 

Friendly Fire Incidents by Type

War                         Air Artillery Ground Antiaircraft

WWII, Korea
& Vietnam               37%   36%        22%        5%
Desert Storm           33%     4%        59%        4%


In the Gulf War, friendly fire by fellow ground troops was easier to report because more of the losses were from long range tank fire and missiles. In both cases, the target wasn't individual soldiers, but the crews of armored vehicles. It's easier to tell if a vehicle, rather than individual, was hit with an American or enemy weapon, and it was more likely that such a detailed examination would be done of destroyed vehicles. In the Gulf War, the official friendly fire losses were 17 percent of our casualties and the fighting in Afghanistan appears to be producing the same pattern. Part of the problem with comparing friendly fire rates from different wars is the change in technology available. The more monitoring you have on your weapons, the easier it is to spot a friendly fire casualty and, of course count it. New weapons also change the types of friendly fire losses. During World War II and Vietnam, misidentification was the cause of (officially counted) friendly fire losses 26 percent of the time. During the Gulf War, when a lot more long range weapons were used, misidentification was the cause of 39 percent of the losses. At the same time, better communications, brought the portion of friendly losses caused by coordination problems down from 45 percent in the 1941-72 period to 25 percent in the Gulf War. 

While it is possible to get friendly fire losses below 20 percent, it isn't going to be easy. Another problem is that American forces tend to have a lot more firepower. This has been a trend that began during World War II and has simply continued since. Also, combat is still a chaotic process and the things don't always work as planned. While the American armed forces are trying to limit the friendly fire losses, they also know that if they impose too many restrictions, the combat troops won't be able to do their job. Even during Vietnam, the communists realized that we were reluctant to use our firepower advantage if the enemy and American troops were too close together. This worked initially, but American troops quickly realized what the communists were trying to do and simply called in the firepower anyway if it appeared that not doing so would lose the battle and get a lot more U.S. troops killed anyway. There is not easy solution to the current problem of friendly fire losses, and probably never will be.